Generative AI-powered search will become the fastest and most efficient way of surfacing information from within large bodies of expert knowledge such as websites, communities, and intranets. Read on to discover how domain experts can make more of their knowledge with GenAI search and how The Digital Twin Hub built a business case.

Hero Image

Search and surface your hidden expertise

The challenge for domain knowledge experts isn’t content creation; it’s retrieving this valuable expertise so it can be used to inform, educate, and help resolve problems. GenAI search will solve this challenge.



Suppose you recall something being published in an article on your website about lowering carbon in Smart Cities. Conventional search will provide links to articles about lowering carbon. However, it is unlikely to surface the reference you need if it is hidden within a document that is not specific to reducing carbon.

Large Language Models (LLMs) like GPT-4 will give you an answer drawn from publicly available online resources but without the source reference. Generic LLMs cannot see beyond paywalls or logins, meaning if the reference you need is on an intranet, forum, or behind a paywall it won’t be included in the answer.

LLM enhanced with RAG (Retrieval Augmented Generation) and semantic search enables the creation of a domain-specific search function. RAG is like an extension pack for LLM, enabling it to draw information from within a specific knowledge base.



Using this type of search, you can find specific references to lowering carbon, whether they exist in an article, presentation, discussion thread, or video. Crucially, it will show you the exact page within a PDF or timestamp in a video where it was mentioned.



Large Language Models (LLMs) like GPT-4 will give you an answer drawn from publicly available online resources but without the source reference. Generic LLMs cannot see beyond paywalls or logins, meaning if the reference you need is on an intranet, forum, or behind a paywall it won’t be included in the answer.

LLM enhanced with RAG (Retrieval Augmented Generation) and semantic search enables the creation of a domain-specific search function. RAG is like an extension pack for LLM, enabling it to draw information from within a specific knowledge base.



Using pioneering AIthis type of search, you can find specific references to lowering carbon, whether they exist in an article, presentation, discussion thread, or video. Crucially, it will show you the exact page within a PDF or timestamp in a video where it was mentioned.

Hero Image

GenAI-powered search collates knowledge from deep within archives, unlocking value from new and historical resources. Here are five ways it will be transformational for domain knowledge experts.

5 ways to make more of your knowledge with LLM

  1. 1. Broaden your portfolio with evergreen subjects


    Published knowledge generally has a shelf-life of only a few months after it has been added to your library and promoted. GenAI search extends the life of this expertise for as long as it remains relevant, creating the opportunity to broaden your content portfolio with evergreen subject hubs.

  2. Protect the integrity of your knowledge


    ChatGPT is trained on publicly available knowledge and reconstituting it as the creator, meaning the original publisher risks losing control and attribution. Furthermore, the integrity of that information will erode over time as it is retrieved and then fed back into ChatGPT. If you have a public-facing knowledge hub and apply your own AI-powered search system, you take ownership with focused search that attributes the source to you and protects the integrity of that information.

  3. Save time with one answer from many sources


    Conventional search returns many answers as complete documents. GenAI search gives you one answer to your question based on many references within your body of knowledge and pinpoints the exact location, resulting in time efficiencies.

  4. Discover what your community is interested in


    Analyse historical and current discussions to understand what’s interesting and useful to your community. Historical search is particularly valuable to domain experts, who have decades of conversations that are still relevant today. 

  5. Drive new leads


    By surfacing historical content, you increase the opportunity to generate new leads without necessarily creating anything new. If knowledge exists behind a paywall, you can still show users a preview and direct them to a registration form to access the complete resource
     

    AI is advancing quickly, but it’s still a relatively new technology that’s not widely understood. We recommend beginning with a Feasibility Audit to determine whether it is viable for your organisation and to identify a low-barrier to-entry Beta test to gain stakeholder buy-in.

Building a LLM business case for Digital Twin Hub

The DT Hub is a community interested in or working in digital twin modelling. Their resources area available to the public but also behind a log-in. They approached us to collate their knowledge and resources via a Wiki, which would have involved many hours of manual work reading and collating resources from hundreds of documents.



We built a beta test to gain buy-in…

Early adopters of GenAI search are the future leaders 
of their field

The DT Hub is ideal for GenAI roll-out, given their extensive knowledge and position as a specialist dedicated to publishing information about digital twins. Once complete, their members will be able to interrogate years of knowledge from the world’s experts in digital twins.

“The board got the idea and were very impressed by the potential of the project and how we're implementing it.
 

 

So Justin and I wanted to thank you for your help in demonstrating that - we definitely got the buy-in we were looking for from board members to proceed with that work!”

Daniel Block

Product Manager

Digital Twin Hub

Luke Wotton

Luke Wotton

Digital Experience Director

If you’re in the top 20% of knowledge providers in your field and you’re not thinking about Generative AI search tools, you risk losing out to competitors already riding the wave. 

I recommend performing an audit on your knowledge to see if its feasible and would make something meaningful for your customers. 

You can download our cheat sheet and do it yourself, or our team can perform it for you and present the results.

work with us